Predicting routes of phase I and II metabolism based on quantum mechanics and machine learning

Unexpected metabolism could lead to the failure of many late-stage drug candidates or even the withdrawal of approved drugs. Thus, it is critical to predict and study the dominant routes of metabolism in the early stages of research.We describe the development and validation of a 'WhichEnzyme' model that accurately predicts the enzyme families most likely to be responsible for a drug-like molecule's metabolism. Furthermore, we combine this model with our previously published regioselectivity models for Cytochromes P450, Aldehyde Oxidases, Flavin-containing Monooxygenases, UDP-glucuronosyltransferases and Sulfotransferases - the most important Phase I and Phase II drug metabolising enzymes - and a 'WhichP450' model that predicts the Cytochrome P450 isoform(s) responsible for a compound's metabolism.The regioselectivity models are based on a mechanistic understanding of these enzymes' actions and use quantum mechanical simulations with machine learning methods to accurately predict sites of metabolism and the resulting metabolites. We train heuristics based on the outputs of the 'WhichEnzyme', 'WhichP450', and regioselectivity models to determine the most likely routes of metabolism and metabolites to be observed experimentally.Finally, we demonstrate that this combination delivers high sensitivity in identifying experimentally reported metabolites and higher precision than other methods for predicting in vivo metabolite profiles.

Structure and Ligand Based Virtual Screening and MPO Topological Analysis of Triazolo Thiadiazepine-fused Coumarin Derivatives as Anti-Parkinson Drug Candidates

Parkinson's disease (PD) is a debilitating condition that can cause locomotor problems in affected patients, such as tremors and body rigidity. PD therapy often includes the use of monoamine oxidase B (MAOB) inhibitors, particularly phenylhalogen compounds and coumarin-based semi-synthetic compounds. The objective of this study was to analyze the structural, pharmacokinetic, and pharmacodynamic profile of a series of Triazolo Thiadiazepine-fused Coumarin Derivatives (TDCDs) against MAOB, in comparison with the inhibitor safinamide. To achieve this goal, we utilized structure-based virtual screening techniques, including target prediction and absorption, distribution, metabolism, and excretion (ADME) prediction based on multi-parameter optimization (MPO) topological analysis, as well as ligand-based virtual screening techniques, such as docking and molecular dynamics. The findings indicate that the TDCDs exhibit structural similarity to other bioactive compounds containing coumarin and MAOB-binding azoles, which are present in the ChEMBL database. The topological analyses suggest that TDCD3 has the best ADME profile, particularly due to the alignment between low lipophilicity and high polarity. The coumarin and triazole portions make a strong contribution to this profile, resulting in a permeability with Papp estimated at 2.15 × 10-5 cm/s, indicating high cell viability. The substance is predicted to be metabolically stable. It is important to note that this is an objective evaluation based on the available data. Molecular docking simulations showed that the ligand has an affinity energy of - 8.075 kcal/mol with MAOB and interacts with biological substrate residues such as Pro102 and Phe103. The results suggest that the compound has a safe profile in relation to the MAOB model, making it a promising active ingredient for the treatment of PD.

Uridine 5'-Diphospho-glucuronosyltransferase 1A3 (UGT1A3) Prediction of Hepatic Clearance of Organic Anion Transporting Polypeptide 1B3 (OATP1B3) Substrate Telmisartan by Glucuronidation Using In Vitro-In Vivo Extrapolation (IVIVE)

BACKGROUND AND OBJECTIVE

The prediction of pharmacokinetic parameters for drugs metabolised by cytochrome P450 enzymes has been the subject of active research for many years, while the application of in vitro-in vivo extrapolation (IVIVE) techniques for non-cytochrome P450 enzymes has not been thoroughly evaluated. There is still no established quantitative method for predicting hepatic clearance of drugs metabolised by uridine 5'-diphospho-glucuronosyltransferases (UGTs), not to mention those which undergo hepatic uptake. The objective of the study was to predict the human hepatic clearance for telmisartan based on in vitro metabolic stability and hepatic uptake results.

METHODS

Telmisartan was examined in liver systems, allowing to estimate intrinsic clearance (CLint, in vitro) based on the substrate disappearance rate with the use of liquid chromatography tandem mass spectrometry (LC-MS/MS) technique. Obtained CLint, in vitro values were corrected for corresponding unbound fractions. Prediction of human hepatic clearance was made from scaled unbound CLint, in vitro data with the use of the well-stirred model, and finally referenced to the literature value of observed clearance in humans, allowing determination of the essential scaling factors.

RESULTS

The in vitro scaled CLint, in vitro by UGT1A3 was assessed using three systems, human hepatocytes, liver microsomes, and recombinant enzymes. Obtained values were scaled and hepatic metabolism clearance was predicted, resulting in significant clearance underprediction. Utilization of the extended clearance concept (ECC) and hepatic uptake improved prediction of hepatic metabolism clearance. The scaling factors for hepatocytes, assessing the in vitro-in vivo difference, changed from sixfold difference to only twofold difference with the application of the ECC.

CONCLUSIONS

The study showed that taking into consideration hepatic uptake of a drug allows us to obtain satisfactory scaling factors, hence enabling the prediction of in vivo hepatic glucuronidation from in vitro data.

Electrochemical transformations catalyzed by cytochrome P450s and peroxidases

Cytochrome P450s (Cyt P450s) and peroxidases are enzymes featuring iron heme cofactors that have wide applicability as biocatalysts in chemical syntheses. Cyt P450s are a family of monooxygenases that oxidize fatty acids, steroids, and xenobiotics, synthesize hormones, and convert drugs and other chemicals to metabolites. Peroxidases are involved in breaking down hydrogen peroxide and can oxidize organic compounds during this process. Both heme-containing enzymes utilize active Fe^IVO intermediates to oxidize reactants. By incorporating these enzymes in stable thin films on electrodes, Cyt P450s and peroxidases can accept electrons from an electrode, albeit by different mechanisms, and catalyze organic transformations in a feasible and cost-effective way. This is an advantageous approach, often called bioelectrocatalysis, compared to their biological pathways in solution that require expensive biochemical reductants such as NADPH or additional enzymes to recycle NADPH for Cyt P450s. Bioelectrocatalysis also serves as an ex situ platform to investigate metabolism of drugs and bio-relevant chemicals. In this paper we review biocatalytic electrochemical reactions using Cyt P450s including C-H activation, S-oxidation, epoxidation, N-hydroxylation, and oxidative N-, and O-dealkylation; as well as reactions catalyzed by peroxidases including synthetically important oxidations of organic compounds. Design aspects of these bioelectrocatalytic reactions are presented and discussed, including enzyme film formation on electrodes, temperature, pH, solvents, and activation of the enzymes. Finally, we discuss challenges and future perspective of these two important bioelectrocatalytic systems.

Predicting Regioselectivity of Cytosolic Sulfotransferase Metabolism for Drugs

Cytosolic sulfotransferases (SULTs) are a family of enzymes responsible for the sulfation of small endogenous and exogenous compounds. SULTs contribute to the conjugation phase of metabolism and share substrates with the uridine 5'-diphospho-glucuronosyltransferase (UGT) family of enzymes. UGTs are considered to be the most important enzymes in the conjugation phase, and SULTs are an auxiliary enzyme system to them. Understanding how the regioselectivity of SULTs differs from that of UGTs is essential from the perspective of developing novel drug candidates. We present a general ligand-based SULT model trained and tested using high-quality experimental regioselectivity data. The current study suggests that, unlike other metabolic enzymes in the modification and conjugation phases, the SULT regioselectivity is not strongly influenced by the activation energy of the rate-limiting step of the catalysis. Instead, the prominent role is played by the substrate binding site of SULT. Thus, the model is trained only on steric and orientation descriptors, which mimic the binding pocket of SULT. The resulting classification model, which predicts whether a site is metabolized, achieved a Cohen's kappa of 0.71.

An overview of compound properties, multiparameter optimization, and computational drug design methods for PARP-1 inhibitor drugs

Breast cancer treatment with PARP-1 inhibitors remains challenging due to emerging toxicities, drug resistance, and unaffordable costs of treatment options. How do we invent strategies to design better anti-cancer drugs? A part of the answer is in optimized compound properties, desirability functions, and modern computational drug design methods that drive selectivity and toxicity and have not been reviewed for PARP-1 inhibitors. Nonetheless, comparisons of these compound properties for PARP-1 inhibitors are not available in the literature. In this review, we analyze the physchem, PKPD space to identify inherent desirability functions characteristic of approved drugs that can be valuable for the design of better candidates. Recent literature utilizing ligand, structure-based drug design strategies and matched molecular pair analysis (MMPA) for the discovery of novel PARP-1 inhibitors are also reviewed. Thus, this perspective provides valuable insights into the medchem and multiparameter optimization of PARP-1 inhibitors that might be useful to other medicinal chemists.

Mechanisms of Electron Transfer Rate Modulations in Cytochrome P450 BM3

Bacterial cytochromes P450 BM3 (CYP450 BM3) catalyze reactions of industrial importance. Despite many successful biotransformations, robust (re)design for novel applications remains challenging. Rational design and evolutionary approaches are not always successful, highlighting a lack of complete understanding of the mechanisms of electron transfer (ET) modulations. Thus, the full potential of CYP450 reactions remains under-exploited. In this work, we report the first molecular dynamics (MD)-based explicit prediction of BM3 ET parameters (reorganization energies; λ and ET free energies; ΔG°), and log ET rates (log k_ET) using the Marcus theory. Overall, the calculated ET rates for the BM3 wild-type (WT), mutants (F393 and L86), ligand-bound state, and ion concentrations agree well with experimental data. In ligand-free (LF) BM3, mutations modulate k_ET via ET ΔG°. Simulations show that the experimental ET rate enhancement is due to increased driving force (more negative ΔG°) upon ligation. This increase is related to the protein reorganization required to accommodate the ligand in the binding pocket rather than binding interactions with the ligand. Our methodology (CYPWare 1.0) automates all the stages of the MD simulation step-up, energy calculations, and estimation of ET parameters. CYPWare 1.0 and this work thus represent an important advancement in the CYP450 ET rate predictions, which has the potential to guide the redesign of ET enzymes. This program and a Web tool are available on GitHub for academic research.

Predicting Regioselectivity of AO, CYP, FMO, and UGT Metabolism Using Quantum Mechanical Simulations and Machine Learning

Unexpected metabolism in modification and conjugation phases can lead to the failure of many late-stage drug candidates or even withdrawal of approved drugs. Thus, it is critical to predict the sites of metabolism (SoM) for enzymes, which interact with drug-like molecules, in the early stages of the research. This study presents methods for predicting the isoform-specific metabolism for human AOs, FMOs, and UGTs and general CYP metabolism for preclinical species. The models use semi-empirical quantum mechanical simulations, validated using experimentally obtained data and DFT calculations, to estimate the reactivity of each SoM in the context of the whole molecule. Ligand-based models, trained and tested using high-quality regioselectivity data, combine the reactivity of the potential SoM with the orientation and steric effects of the binding pockets of the different enzyme isoforms. The resulting models achieve κ values of up to 0.94 and AUC of up to 0.92.

Precision Biotransformation of Emerging Pollutants by Human Cytochrome P450 Using Computational-Experimental Synergy: A Case Study of Tris(1,3-dichloro-2-propyl) Phosphate

Precision biotransformation is an envisioned strategy offering detailed insights into biotransformation pathways in real environmental settings using experimentally guided high-accuracy quantum chemistry. Emerging pollutants, whose metabolites are easily overlooked but may cause idiosyncratic toxicity, are important targets of such a strategy. We demonstrate here that complex metabolic reactions of tris(1,3-dichloro-2-propyl) phosphate (TDCIPP) catalyzed by human CYP450 enzymes can be mapped via a three-step synergy strategy: (i) screening the possible metabolites via high-throughout (moderate-accuracy) computations; (ii) analyzing the proposed metabolites in vitro by human liver microsomes and recombinant human CYP450 enzymes; and (iii) rationalizing the experimental data via precise mechanisms using high-level targeted computations. Through the bilateral dialogues from qualitative to semi-quantitative to quantitative levels, we show how TDCIPP metabolism especially by CYP3A4 generates bis(1,3-dichloro-2-propyl) phosphate (BDCIPP) as an O-dealkylation metabolite and bis(1,3-dichloro-2-propyl) 3-chloro-1-hydroxy-2-propyl phosphate (alcohol_β-dehalogen) as a dehalogenation/reduction metabolite via the initial rate-determining H-abstraction from αC- and βC-positions. The relative yield ratio [dehalogenation/reduction]/[O-dealkylation] is derived from the relative barriers of H-abstraction at the βC- and αC-positions by CYP3A4, estimated as 0.002 to 0.23, viz., an in vitro measured ratio of 0.04. Importantly, alcohol_β-dehalogen formation points to a new mechanism involving successive oxidation and reduction functions of CYP450, with its precursor aldehyde_β-dehalogen being a key intermediate detected by trapping assays and rationalized by computations. We conclude that the proposed three-step synergy strategy may meet the increasing challenge of elucidating biotransformation mechanisms of substantial synthesized organic compounds in the future.

Predicting reactivity to drug metabolism: beyond P450s-modelling FMOs and UGTs

We present a study based on density functional theory calculations to explore the rate limiting steps of product formation for oxidation by Flavin-containing Monooxygenase (FMO) and glucuronidation by the UDP-glucuronosyltransferase (UGT) family of enzymes. FMOs are responsible for the modification phase of metabolism of a wide diversity of drugs, working in conjunction with Cytochrome P450 (CYP) family of enzymes, and UGTs are the most important class of drug conjugation enzymes. Reactivity calculations are important for prediction of metabolism by CYPs and reactivity alone explains around 70-85% of the experimentally observed sites of metabolism within CYP substrates. In the current work we extend this approach to propose model systems which can be used to calculate the activation energies, i.e. reactivity, for the rate-limiting steps for both FMO oxidation and glucuronidation of potential sites of metabolism. These results are validated by comparison with the experimentally observed reaction rates and sites of metabolism, indicating that the presented models are suitable to provide the basis of a reactivity component within generalizable models to predict either FMO or UGT metabolism.

FAME 3: Predicting the Sites of Metabolism in Synthetic Compounds and Natural Products for Phase 1 and Phase 2 Metabolic Enzymes

In this work we present the third generation of FAst MEtabolizer (FAME 3), a collection of extra trees classifiers for the prediction of sites of metabolism (SoMs) in small molecules such as drugs, druglike compounds, natural products, agrochemicals, and cosmetics. FAME 3 was derived from the MetaQSAR database ( Pedretti et al. J. Med. Chem. 2018 , 61 , 1019 ), a recently published data resource on xenobiotic metabolism that contains more than 2100 substrates annotated with more than 6300 experimentally confirmed SoMs related to redox reactions, hydrolysis and other nonredox reactions, and conjugation reactions. In tests with holdout data, FAME 3 models reached competitive performance, with Matthews correlation coefficients (MCCs) ranging from 0.50 for a global model covering phase 1 and phase 2 metabolism, to 0.75 for a focused model for phase 2 metabolism. A model focused on cytochrome P450 metabolism yielded an MCC of 0.57. Results from case studies with several synthetic compounds, natural products, and natural product derivatives demonstrate the agreement between model predictions and literature data even for molecules with structural patterns clearly distinct from those present in the training data. The applicability domains of the individual models were estimated by a new, atom-based distance measure (FAMEscore) that is based on a nearest-neighbor search in the space of atom environments. FAME 3 is available via a public web service at https://nerdd.zbh.uni-hamburg.de/ and as a self-contained Java software package, free for academic and noncommercial research.

Concepts of Artificial Intelligence for Computer-Assisted Drug Discovery

Artificial intelligence (AI), and, in particular, deep learning as a subcategory of AI, provides opportunities for the discovery and development of innovative drugs. Various machine learning approaches have recently (re)emerged, some of which may be considered instances of domain-specific AI which have been successfully employed for drug discovery and design. This review provides a comprehensive portrayal of these machine learning techniques and of their applications in medicinal chemistry. After introducing the basic principles, alongside some application notes, of the various machine learning algorithms, the current state-of-the art of AI-assisted pharmaceutical discovery is discussed, including applications in structure- and ligand-based virtual screening, de novo drug design, physicochemical and pharmacokinetic property prediction, drug repurposing, and related aspects. Finally, several challenges and limitations of the current methods are summarized, with a view to potential future directions for AI-assisted drug discovery and design.

A simple model to solve a complex drug toxicity problem

Linear drug toxicity models like therapeutic index (TI), physicochemical rules (rule of five, 3/75), ligand efficiency indices (LEI), ideal pharmacokinetic (PK) and pharmacodynamic (PD) profiles are widely used in drug discovery and development. In spite of this, predicting drug toxicity at various stages remains challenging and the overall productivity (<20%) and ultimate benefit to the patients remain low. A simple drug toxicity model, "Drug Toxicity Index" (DTI), is developed here using 711 oral drugs. DTI redefines drug toxicity as scaled biphasic and exponential functions of PD, PK and physicochemical parameters. PD, PK and physicochemical toxicity contributions were estimated from the on and off target IC₅₀, maximum unbound plasma drug concentration (free C _max), and log D values, respectively. These contributions are then scaled by molar dose and oral bioavailability and the logarithm of the sum of scaled contributions is DTI. Drugs with DTI above the WHO ATC drug category specific average values consistently have toxic profiles, while drugs with DTI below this average are relatively safe. DTI performs better than standard rules for lead optimization, LEI and exposure based TIs in identifying safe and toxic drugs. DTI classifies 392 drugs reported in the US-FDA's Liver Toxicity Knowledge Base (LTKB) with an AUC for ROC curves of 0.91-0.64 for different WHO ATC categories. DTI has been used to predict network meta-analysis results on relative toxicity within/across eight different therapeutic areas. It is useful in understanding PD, PK and physicochemical toxicity contributions and identifying potentially toxic drugs and the toxicity of recently approved drugs. Decision trees are proposed for applying the DTI concept in preclinical drug discovery and clinical trial settings. DTI can potentially reduce failure in drug discovery and might be useful in therapeutic drug monitoring and in xenobiotic and environmental toxicity studies.

MetScore: Site of Metabolism Prediction Beyond Cytochrome P450 Enzymes

The metabolism of xenobiotics by humans and other organisms is a complex process involving numerous enzymes that catalyze phase I (functionalization) and phase II (conjugation) reactions. Herein we introduce MetScore, a machine learning model that can predict both phase I and phase II reaction sites of drugs in a single prediction run. We developed cheminformatics workflows to filter and process reactions to obtain suitable phase I and phase II data sets for model training. Employing a recently developed molecular representation based on quantum chemical partial charges, we constructed random forest machine learning models for phase I and phase II reactions. After combining these models with our previous cytochrome P450 model and calibrating the combination against Bayer in-house data, we obtained the MetScore model that shows good performance, with Matthews correlation coefficients of 0.61 and 0.76 for diverse phase I and phase II reaction types, respectively. We validated its potential applicability to lead optimization campaigns for a new and independent data set compiled from recent publications. The results of this study demonstrate the usefulness of quantum-chemistry-derived molecular representations for reactivity prediction.